Atomizer/.claude/skills/02_CONTEXT_LOADER.md

skill_id, version, last_updated, type, code_dependencies, requires_skills

skill_id	version	last_updated	type	code_dependencies	requires_skills
SKILL_002	2.0	2025-12-07	loader		SKILL_000

Atomizer Context Loader

Version: 2.0 Updated: 2025-12-07 Purpose: Define what documentation to load based on task type. Ensures LLM sessions have exactly the context needed.

---

Context Loading Philosophy

1. Minimal by default: Don't load everything; load what's needed
2. Expand on demand: Load additional modules when signals detected
3. Single source of truth: Each concept defined in ONE place
4. Layer progression: Bootstrap → Operations → System → Extensions
5. Learn from history: Query LAC for relevant prior knowledge

---

Knowledge Base Query (LAC)

Before starting any task, check LAC for relevant insights:

from knowledge_base.lac import get_lac
lac = get_lac()

# Query relevant insights for the task
insights = lac.get_relevant_insights("bracket mass optimization")

# Check similar past optimizations
similar = lac.query_similar_optimizations("bracket", ["mass"])

# Get method recommendation
rec = lac.get_best_method_for("bracket", n_objectives=1)

Full LAC documentation: .claude/skills/modules/learning-atomizer-core.md

---

Task-Based Loading Rules

CREATE_STUDY

Trigger Keywords: "new", "set up", "create", "optimize", "study"

Always Load:

.claude/skills/core/study-creation-core.md  (SKILL_CORE_001)

Load If:

Condition	Load
User asks about extractors	docs/protocols/system/SYS_12_EXTRACTOR_LIBRARY.md
Telescope/mirror/optics mentioned	modules/zernike-optimization.md
>50 trials OR "neural" OR "surrogate"	docs/protocols/system/SYS_14_NEURAL_ACCELERATION.md
Multi-objective (2+ goals)	docs/protocols/system/SYS_11_MULTI_OBJECTIVE.md
Method selection needed	docs/protocols/system/SYS_15_METHOD_SELECTOR.md

Example Context Stack:

# Simple bracket optimization
core/study-creation-core.md
SYS_12_EXTRACTOR_LIBRARY.md

# Mirror optimization with neural acceleration
core/study-creation-core.md
modules/zernike-optimization.md
SYS_14_NEURAL_ACCELERATION.md
SYS_15_METHOD_SELECTOR.md

---

RUN_OPTIMIZATION

Trigger Keywords: "start", "run", "execute", "begin", "launch"

Always Load:

docs/protocols/operations/OP_02_RUN_OPTIMIZATION.md

Load If:

Condition	Load
"adaptive" OR "characterization"	docs/protocols/system/SYS_10_IMSO.md
"dashboard" OR "real-time"	docs/protocols/system/SYS_13_DASHBOARD_TRACKING.md
"resume" OR "continue"	OP_02 has resume section
Errors occur	docs/protocols/operations/OP_06_TROUBLESHOOT.md

---

MONITOR_PROGRESS

Trigger Keywords: "status", "progress", "how many", "trials", "check"

Always Load:

docs/protocols/operations/OP_03_MONITOR_PROGRESS.md

Load If:

Condition	Load
Dashboard questions	docs/protocols/system/SYS_13_DASHBOARD_TRACKING.md
Pareto/multi-objective	docs/protocols/system/SYS_11_MULTI_OBJECTIVE.md

---

ANALYZE_RESULTS

Trigger Keywords: "results", "best", "compare", "pareto", "report"

Always Load:

docs/protocols/operations/OP_04_ANALYZE_RESULTS.md

Load If:

Condition	Load
Multi-objective/Pareto	docs/protocols/system/SYS_11_MULTI_OBJECTIVE.md
Surrogate accuracy	docs/protocols/system/SYS_14_NEURAL_ACCELERATION.md

---

EXPORT_TRAINING_DATA

Trigger Keywords: "export", "training data", "neural network data"

Always Load:

docs/protocols/operations/OP_05_EXPORT_TRAINING_DATA.md
modules/neural-acceleration.md

---

TROUBLESHOOT

Trigger Keywords: "error", "failed", "not working", "crashed", "help"

Always Load:

docs/protocols/operations/OP_06_TROUBLESHOOT.md

Load If:

Error Type	Load
NX/solve errors	NX solver section of core skill
Extractor errors	modules/extractors-catalog.md
Dashboard errors	docs/protocols/system/SYS_13_DASHBOARD_TRACKING.md
Neural errors	docs/protocols/system/SYS_14_NEURAL_ACCELERATION.md

---

UNDERSTAND_PROTOCOL

Trigger Keywords: "what is", "how does", "explain", "protocol"

Load Based on Topic:

Topic	Load
Protocol 10 / IMSO / adaptive	docs/protocols/system/SYS_10_IMSO.md
Protocol 11 / multi-objective / NSGA	docs/protocols/system/SYS_11_MULTI_OBJECTIVE.md
Extractors / physics extraction	docs/protocols/system/SYS_12_EXTRACTOR_LIBRARY.md
Protocol 13 / dashboard / real-time	docs/protocols/system/SYS_13_DASHBOARD_TRACKING.md
Protocol 14 / neural / surrogate	docs/protocols/system/SYS_14_NEURAL_ACCELERATION.md

---

EXTEND_FUNCTIONALITY

Trigger Keywords: "create extractor", "add hook", "new protocol", "extend"

Requires: Privilege check first (see 00_BOOTSTRAP.md)

Extension Type	Load	Privilege
New extractor	docs/protocols/extensions/EXT_01_CREATE_EXTRACTOR.md	power_user
New hook	docs/protocols/extensions/EXT_02_CREATE_HOOK.md	power_user
New protocol	docs/protocols/extensions/EXT_03_CREATE_PROTOCOL.md	admin
New skill	docs/protocols/extensions/EXT_04_CREATE_SKILL.md	admin

Always Load for Extractors:

modules/nx-docs-lookup.md    # NX API documentation via MCP

---

NX_DEVELOPMENT

Trigger Keywords: "NX Open", "NXOpen", "NX API", "Simcenter", "Nastran card", "NX script"

Always Load:

modules/nx-docs-lookup.md

MCP Tools Available:

Tool	Purpose
siemens_docs_search	Search NX Open, Simcenter, Teamcenter docs
siemens_docs_fetch	Fetch specific documentation page
siemens_auth_status	Check Siemens SSO session status
siemens_login	Re-authenticate if session expired

Use When:

• Building new extractors that use NX Open APIs
• Debugging NX automation errors
• Looking up Nastran card formats
• Finding correct method signatures

---

Signal Detection Patterns

Use these patterns to detect when to load additional modules:

Zernike/Mirror Detection

Signals: "mirror", "telescope", "wavefront", "WFE", "Zernike",
         "RMS", "polishing", "optical", "M1", "surface error"

Action: Load modules/zernike-optimization.md

Neural Acceleration Detection

Signals: "neural", "surrogate", "NN", "machine learning",
         "acceleration", ">50 trials", "fast", "GNN"

Action: Load modules/neural-acceleration.md

Multi-Objective Detection

Signals: Two or more objectives with different goals,
         "pareto", "tradeoff", "NSGA", "multi-objective",
         "minimize X AND maximize Y"

Action: Load SYS_11_MULTI_OBJECTIVE.md

High-Complexity Detection

Signals: >10 design variables, "complex", "many parameters",
         "adaptive", "characterization", "landscape"

Action: Load SYS_10_IMSO.md

NX Open / Simcenter Detection

Signals: "NX Open", "NXOpen", "NX API", "FemPart", "CAE.",
         "Nastran", "CQUAD", "CTRIA", "MAT1", "PSHELL",
         "mesh", "solver", "OP2", "BDF", "Simcenter"

Action: Load modules/nx-docs-lookup.md
        Use MCP tools: siemens_docs_search, siemens_docs_fetch

---

Context Stack Examples

Example 1: Simple Bracket Optimization

User: "Help me optimize my bracket for minimum weight"

Load Stack:
1. core/study-creation-core.md    # Core study creation logic

Example 2: Telescope Mirror with Neural

User: "I need to optimize my M1 mirror's wavefront error with 200 trials"

Load Stack:
1. core/study-creation-core.md         # Core study creation
2. modules/zernike-optimization.md     # Zernike-specific patterns
3. SYS_14_NEURAL_ACCELERATION.md       # Neural acceleration for 200 trials
4. SYS_15_METHOD_SELECTOR.md           # Method recommendation

Example 3: Multi-Objective Structural

User: "Minimize mass AND maximize stiffness for my beam"

Load Stack:
1. core/study-creation-core.md      # Core study creation
2. SYS_11_MULTI_OBJECTIVE.md        # Multi-objective protocol

Example 4: Debug Session

User: "My optimization failed with NX timeout error"

Load Stack:
1. OP_06_TROUBLESHOOT.md            # Troubleshooting guide

Example 5: Create Custom Extractor

User: "I need to extract thermal gradients from my results"

Load Stack:
1. EXT_01_CREATE_EXTRACTOR.md          # Extractor creation guide
2. SYS_12_EXTRACTOR_LIBRARY.md         # Reference existing patterns

---

Loading Priority Order

When multiple modules could apply, load in this order:

1. Core skill (always first for creation tasks)
2. Primary operation protocol (OP_*)
3. Required system protocols (SYS_*)
4. Optional modules (modules/*)
5. Extension protocols (EXT_*) - only if extending

---

Anti-Patterns (Don't Do)

1. Don't load everything: Only load what's needed for the task
2. Don't load extensions for users: Check privilege first
3. Don't skip core skill: For study creation, always load core first
4. Don't mix incompatible protocols: P10 (single-obj) vs P11 (multi-obj)
5. Don't load deprecated docs: Only use docs/protocols/* structure
6. Don't skip LAC query: Always check prior knowledge before starting